|Publication number||US6151329 A|
|Application number||US 08/914,055|
|Publication date||Nov 21, 2000|
|Filing date||Jul 15, 1997|
|Priority date||Jul 17, 1996|
|Also published as||EP0856218A2, WO1998002995A2, WO1998002995A3|
|Publication number||08914055, 914055, US 6151329 A, US 6151329A, US-A-6151329, US6151329 A, US6151329A|
|Inventors||Marouane Berrada, Americo Brajal, Heba Koraitim|
|Original Assignee||U.S. Philips Corporation|
|Export Citation||BiBTeX, EndNote, RefMan|
|Patent Citations (9), Non-Patent Citations (2), Referenced by (45), Classifications (8), Legal Events (5)|
|External Links: USPTO, USPTO Assignment, Espacenet|
1. Field of the Invention
The present invention relates to a method of controlling the transmission of messages between a plurality of stations, said stations being interconnected via a communication channel comprising at least an upstream channel at least partially shared by said stations. The upstream channel is divided into successive frames of predetermined length, each frame consisting of message slots, each of which includes a data part for the transmission of at least a message generated by one of the stations, and M minislots, for sending, in case of unsuccessful transmission of messages, transmission requests in the form of available data slot reservation signals. This method may for instance constitute an appropriate multiple access communication protocol for allowing low level interactivity in satellite networks.
2. Description of the Related Art
In addition to traditional television programmes, it is often proposed to offer to TV users various interactive services and especially the opportunity to exchange information with service providers. Interactivity will be available thanks to a low bitrate return link from user premises to service provider facilities via a geostationary satellite transponder. It is however a challenge to develop a technique that offers a simple and robust channel access to a large number of user terminals with varying traffic volumes, burstiness and time constraints.
A contention-based protocol well adapted to such applications is described for instance in U.S. Pat. No. 4,641,304. As illustrated in FIG. 1 showing an earth station 1 and several frames I, . . . , I+R, I+R+1, of a communication channel 2 shared by a plurality of ground stations (not shown), the earth station or hub 1, which may be one of these stations, constitutes a master station that is able to broadcast via a downstream link a feedback that monitors the upstream link of the channel. In the illustrated case, the round trip delay corresponding to a double hop between the stations and the hub (via a geostationary satellite) is equal to 0,54 second.
As shown in FIG. 2, said protocol, called ARRA for Announced Retransmission Random Access, is based on an organization of the upstream link (or channel) in frames of t(f) seconds in length, composed of K message slots 21(1) to 21(K) and a particular slot called common minislot pool or CMP. Each message slot (or data slot) has a time duration of t(m) and is composed of a data part (or data message DM) and K minislots, and the CMP is also composed of K minislots but without any data part. The time durations of the minislots of the CMP and of each message slot are very small compared to t(m), and the minislots of each message slot occur at the beginning of the message slot.
When a message has to be transmitted from a first station to a second one, this message has to wait until the beginning of the next frame, here called frame I, occurs. Before frame I, all the stations will have received from the hub a feedback message or signal indicating which slot is free for new messages and which is already reserved for messages which had already successfully sent a transmission request. If there are indeed some free message slots, the transmitting station chooses randomly one of these available slots, transmits its message in the data part of this slot, and puts an associated transmission request in one of the minislots (randomly chosen) of the concerned message slot, in order to announce where it will retransmit in case of collision. If there are no available slot, this station puts a transmission request in one of the minislots of the CMP.
After the propagation delay (0,27 second in the above-indicated example), the hub receives all the messages and transmission requests that have been sent in the frame I. It then detects collisions in the data slots and calculates the set of slots that are not reserved (or available slot set ASS) in the frame I+R+1, i.e. in the frame that begins after all the stations have received from the hub the feedback messages (all the stations that have transmitted in frame I being assumed to receive their feedback before frame I+R+1). A slot in said frame I+R+1 is said to be reserved if and only if one request has announced a retransmission in this slot, and on the contrary is available if no request has announced such a retransmission. If more than one request announces a retransmission in the same slot, no successful transmission is possible and this slot is let free for new arrivals, while retransmission requests now occur in the CMP.
After having calculated the available slot set (ASS), i.e. the set of slots available for new messages, the hub broadcasts to all the stations both said ASS and the status of each message slot in the frame (empty, or available=E; used with success=S; collision=C):
(1) if a station has sent a message in a data slot and receives a feedback signal S (successful transmission), it exits the protocol;
(2) if a station receives a feedback signal C (collision) or has sent a request in the CMP, it reads the ASS broadcasted by the hub:
(a) if the reference number of the minislot does not correspond to any of the slots of the ASS, it means that this number corresponds to a reserved slot and that the request for retransmission has been successful: the station sends its message in this reserved data slot, and will then exit the protocol;
(b) if the reference number of the minislot corresponds to a slot of the ASS, it means that the request has failed, and this slot is available for new messages, while the concerned station has to use the CMP for retransmitting its request.
Some computations related to the implementation of this ARRA protocol show that the throughput cannot overpass 0,4 (i.e. 40% of the data slots are not used for transmission). At high loads, more and more requests are indeed sent in the CMP and it becomes impossible to transmit a successful request because of the high probability of collision, in the CMP, between a high number of retransmitted requests trying to use a relatively small amount of minislots.
It is therefore an object of the invention to propose an improved protocol reducing the probability of collisions in the minislots and increasing the performance behaviour of the protocol in terms of average throughput.
To this end the invention relates to a method as defined in the preamble of the description and wherein reservation signals corresponding to retransmission requests take place in anyone of the K×M minislots of the minislot areas of the frame.
The basic idea of said method is to consider all the minislot areas as a large CMP, which allows to reduce the probability of collisions in the minislots. Moreover, by enabling retransmitted requests only in minislots of reserved slots, one avoids that transmission requests of newly arrived messages collide with such retransmitted requests. This solution increases the maximal stable throughput limit to about 0,92 without any noticeably increased complexity.
The present invention will be now described in further detail with reference to the accompanying drawings, in which:
FIG. 1 shows schematically a communication channel divided into successive frames and a master station provided for monitoring the transmission of messages between stations via said communication channel;
FIGS. 2 and 3 illustrate the structure of each frame when using respectively the known protocol ARRA and the protocol according to the invention; and
FIG. 4 illustrates in its upper part (references 41 to 49) the procedure executed in a station until the status of a transmitted message is S and in its lower part (references 51 to 55) the procedure in the hub.
In fact it has been seen that the limitation of the ARRA protocol is caused by the high number of requests colliding in the CMP, said requests having very little chance to be accepted by the hub since they frequently collide in a small number of minislots. According to the invention, it has been found that the probability of collision is reduced if all the minislot areas of the frame are considered as a larger CMP. The upstream link now comprises frames including K message slots and each of these slots contains, as indicated in FIG. 3, a data slot and M minislots, the total number of minislots being K×M. The feedback is transmitted (downlink) in a frame of K slots, each of them comprising two pieces of information:
(a) a first one (2 bits) for the status E (empty), S (successful) or C (collision) of the corresponding data slot in the upstream frame;
(b) a second one (at most 6 bits in this case) for the identification of the winning minislot, i.e. of the minislot really used by the station that is allowed to transmit a message in the data part of the corresponding slot in the next upstream frame (said information is equal to a number between 1 and K2, and equal to 0 if the data part of said message slot is let free for newly arrived messages).
The procedure executed in a station until the status of a transmitted message is S is now described with reference to FIG. 4. When new messages NM, available at the output of a message generator 41, arrive at a station during a frame I-1, they have to wait for the following frame I (before the beginning of this frame I, they receive the downstream frame containing the feedback sent by the hub). A first test 42 is then carried out:
(a) if one or some slots are free (output Y), the station (step 43) chooses randomly one of them (D=number of said free slot), transmits the message in this slot, and transmits a request in one of the M minislots of said slot D, while keeping in memory a copy of the message and the minislot number d;
(b) if no slot is free for new arrivals (output N), the concerned station (step 44) keeps a copy of the message and transmits a reservation signal (=message transmission request) in one of the K×M minislots of the frame, while keeping in memory the number J of said minislot.
After the trip delay, the downstream frame corresponding to upstream frame I is received (step 45) by the station. A second test 46 is then carried out:
(a) if a message has been transmitted in slot D of upstream frame I (output N, corresponding to the situation "no collision"), the station looks at the feedback contained in slot D of the received downstream frame: if the status is S (successful transmission), the copy of the message is deleted (step 47);
(b) if the status of the data slot is C (collision, output Y) or no message has been transmitted (only a request), the station looks at the feedback contained in the slot corresponding to J (the number of the selected minislot during the step 43) and a third test 48 is carried out:
(i) if the identification of the winning slot is J, it means that this slot is reserved for this station in the upstream frame (I+R+1): the station will transmit its message in said reserved slot (step 49), without any request since it would be useless;
(ii) if the identification of the winning slot is not J, the station cannot transmit its message: it transmits a reservation signal (step 44) in one of the minislots of the reserved slots and keeps in memory the corresponding minislot number.
In the hub, the procedure, also illustrated in FIG. 4, is the following. When the hub receives the frame I (step 51), it collects (detection step 52) all the minislots where successful requests have been transmitted and all the data slots with successful message transmission (the other data slots and minislots of frame I corresponding to detected collisions). Among these data slots, some of them have been sent as new arrivals in a free slot of the frame I, and the corresponding request for retransmission must be discarded from the number Jk of said minislots with successful requests. The resulting set of data slots is Sk. For the discarding step, it is needed that the hub keeps in memory the status of the slots in frame I (free for new arrivals, or reserved), and (R+1) frames status must therefore be memorized (step 53) by the hub in order to calculate Sk. The feedback formulation then takes place: the requested slots are allocated to respective reservations signals (step 54), each concerned slot being then reserved, and, as already seen, the hub puts two pieces of information in each slot of the downstream frame (the status E, S or C of the data slot, and the number Jk identifying the winning minislot, i.e. the number of the reservation signal corresponding to the allowed transmission in said data slot in the next upstream frame), said downstream frame broadcasting being the step 55 of the hub procedure.
The invention is obviously not limited to the previous implementation, from which other ones may be deduced without departing from the scope of the invention. For instance, the communication method according to the invention may have some jitter at high load. In order to reduce it, while the successful requests (i.e. those which will have a reserved slot in the next frame) are chosen by the hub among those which have not collided in the minislots of the current frame, the hub may take these requests into account more efficiently by putting them in a FIFO queue. The access delay is then composed now of three components:
(a) waiting for the next frame delay, i.e. the time period between the arrival of the message and the beginning of the frame in which the message is transmitted for the first time;
(b) contention delay, i.e. the time period between the beginning of the frame in which the message is transmitted for the first time and the complete reception of the frame in which a request is received successfully in the hub;
(c) waiting for transmission delay, i.e. the time period between the complete reception of the frame in which a request is received successfully in the hub and the first slot in which the message is transmitted.
The introduction of said transmission queue adds the waiting for transmission delay component but reduces the contention delay, which reduces noticeably the jitter because said contention delay is a multiple of the round trip time (=n×0,54 sec. in the above-indicated example), while the waiting for transmission delay can be less than one frame (i.e. a few slots). The jitter is reduced even at high loads since the majority of the packets are transmitted after a first successful transmission. They suffer a given jitter due to the queuing delay but which is much more smaller than the jitter caused by the retransmission of a request. Nevertheless the packets that suffer a collision in the data slot and in the minislot (which is quite rare) have to retransmit a request. In order to reduce the queuing delay for these old contentions and to avoid a competition with new arrived messages, a priority in the transmission queue may be given to them.
In all the cases, the proposed method is a very good candidate for instance for satellite channel protocols with a very large bursty population, especially for applications involving the generation of short constant length messages (interactive applications such as authentication, pay-per-view, data base consultations, but also all the applications that require the establishment of a connection within a short time).
|Cited Patent||Filing date||Publication date||Applicant||Title|
|US4641304 *||Jun 6, 1986||Feb 3, 1987||Rca Corporation||Announced retransmission random access system|
|US5185737 *||Mar 12, 1990||Feb 9, 1993||International Business Machines Corporation||Method and apparatus for cyclic reservation multiple access in a communications system|
|US5278833 *||Mar 28, 1991||Jan 11, 1994||Motorola, Inc.||Method for providing reserved communication access using multiple random access resources|
|US5303234 *||Jul 15, 1992||Apr 12, 1994||Nec Corporation||Random access data communication system with slot assignment capability for contending users|
|US5390181 *||Jun 4, 1993||Feb 14, 1995||Illinois Institute Of Technology||Method for detecting collisions on and controlling access to a transmission channel|
|US5404353 *||May 12, 1994||Apr 4, 1995||Digital Equipment Corp.||Dynamic defer technique for traffic congestion control in a communication network bridge device|
|US5590131 *||May 30, 1995||Dec 31, 1996||Motorola, Inc.||Efficient distributed queueing random access method for the medium access control layer in networks with broadcast channels|
|US5896385 *||Aug 6, 1996||Apr 20, 1999||U.S. Philips Corporation||Transmission control method between a plurality of stations and corresponding stations and communication systems|
|US5917813 *||Oct 23, 1996||Jun 29, 1999||U.S. Philips Corporation||Multiple access telecommunication network|
|1||*||Announced Retransmission Random Access Protocoles, Dipankar Raychaudhuri, IEEE Transactions on Communications, vol. COM 33, No. 11, Nov. 1985.|
|2||Announced Retransmission Random Access Protocoles, Dipankar Raychaudhuri, IEEE Transactions on Communications, vol. COM-33, No. 11, Nov. 1985.|
|Citing Patent||Filing date||Publication date||Applicant||Title|
|US6249677 *||Feb 11, 1999||Jun 19, 2001||Hughes Electronics Corporation||Apparatus and method for delivering key information of a channel request message from a user terminal to a network|
|US6826589 *||Nov 30, 1999||Nov 30, 2004||Koninklijke Philips Electronics N.V.||Request mechanism for a communication system between interactive terminals and master station|
|US6847626 *||Sep 28, 1999||Jan 25, 2005||Tachyon, Inc.||Method and apparatus for multiple access in a communication system|
|US6928065 *||Jun 11, 2002||Aug 9, 2005||Motorola, Inc.||Methods of addressing and signaling a plurality of subscriber units in a single slot|
|US7203204 *||Jan 24, 2005||Apr 10, 2007||Tachyon Networks Incorporated||Method and apparatus for multiple access in a communication system|
|US7227871 *||Jul 31, 2002||Jun 5, 2007||Broadcom Corporation||Method and system for real-time change of slot duration|
|US7403538||Apr 21, 2004||Jul 22, 2008||At&T Corporation||Voice data integrated multiaccess by self-reservation and contention algorithm|
|US7420950 *||Apr 5, 2004||Sep 2, 2008||Dspg Ltd.||Multiplex communication with slotted retransmission on demand|
|US7436809 *||Jan 6, 2005||Oct 14, 2008||National Institute Of Information And Communications Technology, Incorporated Administrative Agency||Communicating system, communicating method, base station, and mobile station|
|US7630351||Sep 29, 2005||Dec 8, 2009||At&T Corp.||RSVP/SBM based up-stream session setup, modification, and teardown for QoS-driven wireless LANs|
|US7639657||Jan 17, 2006||Dec 29, 2009||At&T Corp.||Centralized contention and reservation request for QoS-driven wireless LANs|
|US7646756||Jan 12, 2010||At&T Corp.||Multipoll for QoS-driven wireless LANs|
|US7664068 *||Jul 11, 2008||Feb 16, 2010||At&T Corp.||Voice data integrated multiaccess by self-reservation and contention algorithm|
|US7738378||Apr 28, 2005||Jun 15, 2010||At&T Intellectual Property Ii, L.P.||RSVP/SBM based side-stream session setup, modification, and teardown for QoS-driven wireless LANs|
|US7756092||Jul 18, 2005||Jul 13, 2010||At&T Intellectual Property Ii, L.P.||In-band QoS signaling reference model for QoS-driven wireless LANs connected to one or more networks|
|US7756095||Oct 31, 2007||Jul 13, 2010||At&T Intellectual Property Ii, L.P.||In-band QoS signaling reference model for QoS-driven wireless LANs|
|US7852870||Oct 5, 2006||Dec 14, 2010||Broadcom Corporation||Method for real-time change of slot duration within a wireless network|
|US7852871||Oct 5, 2006||Dec 14, 2010||Broadcom Corporation||System for real-time change of slot duration within a wireless network|
|US7860053||Jul 18, 2005||Dec 28, 2010||At&T Intellectual Property Ii, L.P.||Voice-data integrated multiaccess by self-reservation and stabilized aloha contention|
|US7870195 *||Jan 24, 2001||Jan 11, 2011||Emc Corporation||Inter-processor messaging|
|US7899012||Dec 27, 2009||Mar 1, 2011||At&T Intellectual Property Ii, L.P.||Virtual streams for QOS-driven wireless LANS|
|US7986951||Dec 10, 2008||Jul 26, 2011||Inmarsat Global Limited||Method and apparatus for controlling transmissions in communication systems|
|US8009649||Aug 30, 2011||At&T Intellectual Property Ii, L.P.||Admission control for QoS-driven wireless LANs|
|US8014372||Dec 8, 2009||Sep 6, 2011||At&T Intellectual Property Ii, L.P.||Multipoll for QoS-driven wireless LANs|
|US8130732||Mar 14, 2006||Mar 6, 2012||At&T Intellectual Property Ii, L.P.||Enhanced channel access mechanisms for QoS-driven wireless LANs|
|US8149812 *||Sep 15, 2008||Apr 3, 2012||Samsumg Electronics Co., Ltd.||Method and apparatus for retransmitting data in wireless LAN system using visible light communication|
|US8320355 *||Oct 23, 2007||Nov 27, 2012||At&T Intellectual Property Ii, L.P.||Voice data integrated multiaccess by self-reservation and contention algorithm|
|US8437323||May 7, 2013||At&T Intellectual Property Ii, L.P.||Admission control for QoS-driven wireless LANs|
|US8503414||Dec 4, 2009||Aug 6, 2013||At&T Intellectual Property Ii, L.P.||RSVP/SBM based up-stream session setup, modification, and teardown for QoS-driven wireless LANs|
|US8576827||Sep 14, 2012||Nov 5, 2013||At&T Intellectual Property Ii, L.P.||Voice data integrated multiaccess by self-reservation and contention algorithm|
|US8605707||Mar 5, 2012||Dec 10, 2013||At&T Intellectual Property Ii, L.P.||Enhanced channel access mechanisms for QoS-driven wireless LANs|
|US8811165||Dec 27, 2009||Aug 19, 2014||At&T Intellectual Property Ii, L.P.||Voice-data integrated multiaccess by self-reservation and stabilized aloha contention|
|US8855060||Dec 8, 2009||Oct 7, 2014||At&T Intellectual Property Ii, L.P.||Centralized contention and reservation request for QoS-driven wireless LANs|
|US8892706 *||Jun 21, 2010||Nov 18, 2014||Vmware, Inc.||Private ethernet overlay networks over a shared ethernet in a virtual environment|
|US8989165||May 6, 2013||Mar 24, 2015||At&T Intellectual Property Ii, L.P.||Admission control for QoS-driven wireless LANs|
|US20020075891 *||Dec 16, 2000||Jun 20, 2002||Slim Souissi||Network assisted random access method|
|US20030058892 *||Jul 31, 2002||Mar 27, 2003||Broadcom Corporation||Method and system for real-time change of slot duration|
|US20030187981 *||Nov 30, 1999||Oct 2, 2003||Marouane Berrada||Request mechanism for a communication system between interactive terminals and master station|
|US20030227898 *||Jun 11, 2002||Dec 11, 2003||Logalbo Robert D.||Methods of addressing and signaling a plurality of subscriber units in a single slot|
|US20050149649 *||Jan 24, 2005||Jul 7, 2005||Carneal Bruce L.||Method and apparatus for multiple access in a communication system|
|US20050226198 *||Apr 5, 2004||Oct 13, 2005||Barak Ilan S||Multiplex communication with slotted retransmission on demand|
|US20070004437 *||Jan 6, 2005||Jan 4, 2007||Hiroshi Harada||Communicating system, communicating method, base station, and mobile station|
|US20070064741 *||Oct 5, 2006||Mar 22, 2007||Broadcom Corporation||Method for real-time change of slot duration within a wireless network|
|US20090073952 *||Sep 15, 2008||Mar 19, 2009||Hong-Seok Shin||Method and apparatus for retransmitting data in wireless lan system using visible light communication|
|US20090097435 *||Dec 10, 2008||Apr 16, 2009||Inmarsat Ltd.||Method and apparatus for controlling transmissions in communication systems|
|U.S. Classification||370/447, 370/348|
|International Classification||H04J3/16, H04B7/212, H04J3/00, H04L12/56|
|Jan 9, 1998||AS||Assignment|
Owner name: U.S. PHILIPS CORPORATION, NEW YORK
Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:BERRADA, MAROUANE;BRAJAL, AMERICO;KORAITIM, HEBA;REEL/FRAME:008891/0392;SIGNING DATES FROM 19971106 TO 19971110
|Apr 26, 2004||FPAY||Fee payment|
Year of fee payment: 4
|Jun 2, 2008||REMI||Maintenance fee reminder mailed|
|Nov 21, 2008||LAPS||Lapse for failure to pay maintenance fees|
|Jan 13, 2009||FP||Expired due to failure to pay maintenance fee|
Effective date: 20081121